Wednesday, 3 October 2007

SPC

According to Wikipedia, lightning can travel at a speed of 100,000 MPH, however I think storage vendors are even faster than lightning when it comes to highlighting or dissing the competition.

Mere microseconds after reading Claus Mikkelsen's blog on the USP-V SPC figures, there are posts from BarryW and BarryB, doing the highlighting and dissing respectively (I almost wrote respectfully there; that would have been a funny typo). BarryB must have no real work to do other than to write his blog, looking at the size of the posts he does!

Anyway. I'm not going to comment on the results because the others have done that enough already and I don't think the details are that relevant. I think what's more relevant is the stance EMC are taking in not providing figures for customers on the performance of their equipment. I can't decide whether its a case of arrogance and therefore a feeling they don't need to provide details because as BarryB says, the customer will buy anyway, or is it because the DMX will not match up to the performance of its competitors. I think it is a mixture of both.

EMC aren't an array vendor any more and haven't been for a long time. OK, it is the product they're most remembered for historically, but their reach is now so wide and deep I think Symmetrix isn't the focus of a lot of their attentions. If it was, DMX4 would not just scale by the GB, it would have more connectivity, more cache and EMC would have been the *leader* in the implementation of technology like thin provisioning, not the also ran.

On reflection, I think EMC should provide SPC figures. If DMX is better than the others and is "Simply the Best" prove it; bragging starts to sound hollow after a while.

3 comments:

Mark said...

EMC does provide accurate performance numbers to customers but does so specifically for the workloads they plan to run on the systems. What it doesn't do is run a benchmark designed solely for the purposes of issuing a press release.

That's what SPC is, a press release benchmark. When it became clear that that was all it was EMC bailed on it. A lot of other vendors did too, HDS included, until they finally got the press release they wanted out of it.

We've seen this before with the processor wars and the multi-processor system wars where every six months a CPU or systems vendor would fire off a new press release claiming that they were number one while providing sets of numbers which became so incredibly twisted that by the end they looked like balloon animals.

EMC has trained & rigorously tested Speed Champions for the different platforms in every region. It has also published ESRP docs (Which are not benchmarks but will give accurate performance planning information using best practices specific to the application workload), and is in the process of doing the same thing for other application workloads.

I don't get the comfort factor of SPC. Essentially the argument here is why don't we test arrays the way we do 3D graphics cards in PC magazines. But when they're testing graphics cards at least they use games people will run on those cards to do so. You won't see an SPC style workload unless I finish building that time machine and we go back to 1995. It's a technically bogus test.

As for being an also ran Celerra thin provisioning shipped in 2006 for NAS & iSCSI environments. Why there first? Because that was the market segment which demanded it first. It's already been announced that the thin provisioning will be a feature of FC storage in 08.

That's all it is after all. A feature.

hollis_chuck said...

Hi Chris

By your post, I assume that you believe that the SPC in its current form is a useful and valid test for comparison of storage arrays.

That's what has got me stumped. It's a test alright, but is it a useful test?

I don't think you'd be promoting its use unless you thought its result somehow mapped into the real world.

Many of us can't find that linkage. We know SPC measures something, but what does it really measure?

I have my thoughts -- what do you think?

Chris M Evans said...

Chuck

Personally I think we need *something* that gives us an idea of a benchmark, even if it is not perfect. Things never will be perfect because each vendor must design their products uniquely with significant differences otherwise they'd be suing each other all the time!

Those of us who've installed DMX, Shark and USP know what they can do, however when EMC claims one generation of a product provides x% improvement over the previous, it helps to have independent testing to prove it. I think it also helps to have some measure of comparison, even if it is flawed.